Refining the method


 

In a comment, I lately described the four stages of my method.

  • Unearth the logic. That takes quite some fiddling around.
  • Transform the logic into patterns that are more or less position independent
  • Remember the patterns
  • Test the absorption
This method works to a certain degree. With a pace of one problem per day and, say, 3 patterns per problem. about 1000 new patterns per year seems to be doable. But it is hard work, and although the method works,  I have somehow the feeling that I haven't reached the maximum learning potential of the method.

The method is pretty crude, so let's have a look whether it is possible to refine it.

Unearth the logic
In this area, you might be able to snatch a free lunch or two. Don't waste your time with trying to solve the problems yourself. Use Stockfish and fiddle around as much as you can. Make use of every annotation by grandmasters if you can. Use themed problem sets whenever possible.

Since this is a system 2 activity, it will take a lot of time by definition. Do not skimp here. Because you conceptualize here, you make on solution available to many positions. Go to the extremes here.

Transform the logic into patterns
At the end of the day, you need to absorb patterns. How do these patterns look like? Let me give an example.

White to move

r1b1r1k1/pq3p1p/1p2np1Q/5N2/3Rp3/6PB/P4P1P/6K1 w - - 0 1

Sandipan, Chanda vs. Tiviakov, Sergei (2007)

Let's have a look at a few patterns. In no particular order.

Pattern 1

Pattern 1 that emerges is the killzone. The killzone is the dynamic killbox. Rd4 radiates one wall of the killzone. This inspires moves that are trying to create a hole in that wall, through which the black king might escape.

Pattern 2

Bd7 shields an escape route for the black king. It comes with a downside though.

Pattern 3

Bd7 shields the black Queen from protecting f7.

Pattern 4

Pattern 4 is based on the quiet move Qxf6, which is designed to clear the square h6 for the white knight. You need to have already a picture of the killzone and the king chase in mind.

Remember the patterns
The Movetrainer of Chessable is a good help in presenting you with the right patterns at the right time according the spaced repetition principle. Beware though that you don't use it to remember the moves but the patterns.

Test the absorption
The stopwatch is the main tool to proof that you have absorbed the patterns. In practice you will not use a stopwatch probably. You will just know when you have absorbed the patterns because it feels different.

The sentence metaphor
The patterns form the words of which you can make a logical sentence. The quiet move of pattern 4 can only be absorbed in relation to the sequel. You have to understand and absorb the whole sentence.

Where can we improve?
I noticed 1.5 year ago that I'm bad at knight moves.Since then I trained with a problem set of 209 knight problems. The position in this post is from that problem set. 

I still happen to fall for a knight fork every now and then. I realized that I have absorbed the moves, but not so much the patterns. Hence I don't need to look for more problems, but for a better absorption.

This post is an attempt to improve on stage 1. It is important to get the patterns straight. To have a clear picture of the killzone, and how the knight and the queen combine in chasing the black king.

Furthermore, I belief that stage 3 needs some improvement. It is tempting to absorb the knight moves instead of the patterns. I'm experimenting with more visualisation in this stage.

Visualization
We often talk about seeing things and visualization. But that might not be the best way to formulate matters. Indeed, I'm rather visual oriented, and I belief that Robert is too. Yet people can differ very much in the way how they represent a pattern internally. Some people are more verbally oriented, while others suffer from aphantasia. So I guess representation is more neutral than visualization.









Comments

  1. You are correct: I am a visually oriented person.

    "SEEING" is similar (by analogy) to visualization but, IMHO, is distinctly different. I think of "SEEING" as (in essence) the Gestault phenomenon: an organized whole that is perceived as more than the sum of its parts. In the present context, that means "SEEING" the words and the sentence in terms of its meaning within a larger context, without focusing on any particular word, and comprehending that meaning in its entirety.

    It's the difference between saying something meaningful and spewing a word salad.

    ReplyDelete
  2. FEN = r4rk1/1pp2ppp/p7/3p4/B3n2q/5N2/PPP3K1/RN1Q2R1 b - - 0 18

    The problem position is simple and and the solution is obvious. However, the game itself is a good illustration of cracking open the killzone and pursuing the enemy king regardless of the cost.

    Puzzle #T8ige
    Rating: 1134
    Played 453 times

    [Event "rated blitz game"]
    [Site "https://lichess.org/QXki4Ba6"]
    [Date "2024.05.18"]
    [White "Gonzo_DO"]
    [Black "KimuraTrap"]
    [Result "0-1"]
    [GameId "QXki4Ba6"]
    [UTCDate "2024.05.18"]
    [UTCTime "07:06:51"]
    [WhiteElo "2121"]
    [BlackElo "2205"]
    [WhiteRatingDiff "-5"]
    [BlackRatingDiff "+4"]
    [Variant "Standard"]
    [TimeControl "180+2"]
    [ECO "C80"]
    [Opening "Ruy Lopez: Open, Riga Variation"]
    [Termination "Normal"]
    [Annotator "lichess.org"]

    1. e4 { [%eval 0.14] [%clk 0:03:00] } 1... e5 { [%eval 0.21] [%clk 0:03:00] } 2. Nf3 { [%eval 0.19] [%clk 0:03:00] } 2... Nc6 { [%eval 0.19] [%clk 0:03:00] } 3. Bb5 { [%eval 0.18] [%clk 0:03:00] } 3... a6 { [%eval 0.38] [%clk 0:03:01] } 4. Ba4 { [%eval 0.21] [%clk 0:03:01] } 4... Nf6 { [%eval 0.25] [%clk 0:03:00] } 5. O-O { [%eval 0.29] [%clk 0:03:01] } 5... Nxe4 { [%eval 0.27] [%clk 0:02:58] } 6. d4 { [%eval 0.31] [%clk 0:03:01] } 6... exd4?! { (0.31 -1.94) Blunder. Rxe4+ was best. } { [%eval -1.94] [%clk 0:02:43] } (11. Rxe4+ dxe4 12. Qd8+ Qxd8 13. Nxd8+ Kxd8 14. Kxh2 Be6 15. Bf4 Kc8 16. Nd2 f5) 11... Qxg5 { [%eval -1.96] [%clk 0:02:11] } 12. Kxh2?! { (-1.96 -1.25) Mistake. Qh4+ was best. } { [%eval -1.25] [%clk 0:01:50] } (12... Qh4+ 13. Kg1 Qxf2+ 14. Kh2 Qh4+ 15. Kg1 O-O 16. Nc3 Nxc3 17. bxc3 Qxa4 18. Ne7+) 13. Nd4?? { (-1.25 Mate in 3) Checkmate is now unavoidable. Qe1 was best. } { [%eval #-3] [%clk 0:02:11] } (17. Qe1 Nf2+ 18. Qxf2 Qxf2 19. gxh3 Qxd4 20. Nc3 c6 21. Rgf1 Qh4 22. Rf3 b5) 17... Bxg2+ { [%eval #-2] [%clk 0:00:40] } 18. Kxg2 { [%eval #-2] [%clk 0:02:11] } 18... Qg3+ { [%eval #-1] [%clk 0:00:33] } 19. Kh1 { [%eval #-1] [%clk 0:02:12] } 19... Nf2# { [%clk 0:00:28] } { Black wins by checkmate. } 0-1

    ReplyDelete
  3. I consider kids to become grandmaster at age 14 or 15 as a proof that it is based on a trick. How far can this trick bring you?

    "Trick" doesn't sound very respectful maybe. But that is not what I mean. It is by no means an easy trick that everybody can learn on a rainy day.

    Compare it wit learning a language. How far can you come with the "easy" trick of just learning words?

    Learning a language says nothing about the story you have to tell. The study of master games is just that. Studying stories instead of sentences or words.

    Words can bring you to candidat master. Sentences brings you to grandmaster. Being able to tell a story is what you need to become a super grandmaster.

    I have seen a lot of videos the past two years, and a lot of commenters on high level tournaments.

    I saw a blitz game between Magnus Carlson and a grandmaster. What I saw was a superior ability to formulate chess sentences.

    Chess sentences are the scenarios in my system. And learning scenarios is the essence of the "trick".

    ReplyDelete

Post a Comment

Chessbase PGN viewer